SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2210.09622 · 2022

Deep Black-Box Reinforcement Learning with Movement Primitives

arXiv · PDF · Open in the Atlas

Code that ran

5 of 6 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2210.09622.json · badge: /api/ran/2210.09622.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
ALRhub/fancy_gym canonical (classifier label)
classifier basis not recorded on the link
5 ran · 1 no recorded run
FunctionStatusWhere it lives
angle_normalize Ran
ran on a synthesized input
ALRhub/fancy_gym/fancy_gym/utils/utils.py
code served (permissive licence) · get_code("38fb46c86c62915f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (4,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Test case 1: angle in degrees
x_deg = np.array([45.0, 180.0, -270.0, 720.5])
type_deg = 'deg'

# Test case 2: angle in radians
x_rad = np.array([np.pi/4, np.pi, -3*np.pi/2, 4*np.pi])
type_rad = 'rad'

# Use the degree case as the primary test
ARGS = {
    'x': x_deg,
    'type': type_deg
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (4,) — shape (4,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
ensure_finite_time Ran
ran on a synthesized input
ALRhub/fancy_gym/fancy_gym/utils/make_env_helpers.py
code served (permissive licence) · get_code("f6d6464f41ab5319")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned MinimalEnv. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import gymnasium as gym
from gymnasium.wrappers import TimeLimit
import numpy as np

# Create a minimal gym environment
class MinimalEnv(gym.Env):
    def __init__(self, max_steps=None):
        super().__init__()
        self.observation_space = gym.spaces.Box(low=0, high=1, shape=(4,), dtype=np.float32)
        self.action_space = gym.spaces.Discrete(2)
        self._max_steps = max_steps
        self.spec = type('obj', (object,), {'max_episode_steps': max_steps})()
    
    def reset(self, seed=None):
        return self.observation_space.sample(), {}
    
    def step(self, action):
        obs = self.observation_space.sample()
        reward = 0.0
        terminated = False
        truncated = False
        info = {}
        return obs, reward, terminated, truncated, info
    
    def render(self):
        pass

class MinimalEnvWithMaxPathLength(gym.Env):
    def __init__(self):
        super().__init__()
        self.observation_space = gym.spaces.Box(low=0, high=1, shape=(4,), dtype=np.float32)
        self.action_space = gym.spaces.Discrete(2)
        self.max_path_length = 300
        self.spec = type('obj', (object,), {'max_episode_steps': None})()
    
    def reset(self, seed=None):
        return self.observation_space.sample(), {}
    
    def step(self, action):
        obs = self.observation_space.sample()
        reward = 0.0
        terminated = False
        truncated = False
        info = {}
        return obs, reward, terminated, truncated, info
    
    def render(self):
        pass

# Create test environments
env_with_limit = MinimalEnv(max_steps=200)
env_without_limit = MinimalEnv(max_steps=None)
env_with_max_path_length = MinimalEnvWithMaxPathLength()

ARGS = {
    'env': env_with_limit,
    'fallback_max_steps': 500
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
gymnasium, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type MinimalEnv — a MinimalEnv output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MinimalEnv output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_numpy Ran
ran on a synthesized input
ALRhub/fancy_gym/fancy_gym/utils/utils.py
code served (permissive licence) · get_code("f4490122e8103845")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch as ch
import torch.nn as nn

ch.manual_seed(0)

# Create a small tensor to pass to get_numpy
# The function calls .detach().cpu().numpy() on it
x = ch.randn(2, 4)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
nested_update Ran
ran on a synthesized input
ALRhub/fancy_gym/fancy_gym/envs/registry.py
code served (permissive licence) · get_code("adfe7e4509baf685")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from collections.abc import MutableMapping

torch.manual_seed(0)

# Create a simple nested mutable mapping for base
base = {
    'model': {
        'layers': 2,
        'hidden_dim': 64
    },
    'training': {
        'lr': 0.001,
        'epochs': 10
    },
    'data': {
        'batch_size': 32
    }
}

# Create an update mapping with nested structure
update = {
    'training': {
        'lr': 0.0001,
        'optimizer': 'adam'
    },
    'data': {
        'batch_size': 16,
        'shuffle': True
    }
}

ARGS = {
    'base': base,
    'update': update
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
nested_update Ran
ran on a synthesized input
ALRhub/fancy_gym/fancy_gym/utils/utils.py
code served (permissive licence) · get_code("893b128289093642")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from collections.abc import Mapping, MutableMapping

torch.manual_seed(0)

# Construct nested base mapping
base = {
    "model": {
        "learning_rate": 0.001,
        "layers": 3,
        "activation": "relu"
    },
    "training": {
        "batch_size": 32,
        "epochs": 10
    },
    "debug": False
}

# Construct nested update mapping with overlapping and new keys
update = {
    "model": {
        "learning_rate": 0.002,
        "dropout": 0.5
    },
    "training": {
        "epochs": 20
    },
    "validation": {
        "split": 0.2,
        "metric": "accuracy"
    }
}

ARGS = {
    "base": base,
    "update": update
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_env_duration Not yet run
unverified (harvested, no recorded run)
ALRhub/fancy_gym/fancy_gym/utils/make_env_helpers.py
code served (permissive licence) · get_code("1e0e20844a795e50")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 6 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

\Episode-based reinforcement learning (ERL) algorithms treat reinforcement learning (RL) as a black-box optimization problem where we learn to select a parameter vector of a controller, often represented as a movement primitive, for a given task descriptor called a context. ERL offers several distinct benefits in comparison to step-based RL. It generates smooth control trajectories, can handle non-Markovian reward definitions, and the resulting exploration in parameter space is well suited for solving sparse reward settings. Yet, the high dimensionality of the movement primitive parameters has so far hampered the effective use of deep RL methods. In this paper, we present a new algorithm for deep ERL. It is based on differentiable trust region layers, a successful on-policy deep RL algorithm. These layers allow us to specify trust regions for the policy update that are solved exactly for each state using convex optimization, which enables policies learning with the high precision required for the ERL. We compare our ERL algorithm to state-of-the-art step-based algorithms in many complex simulated robotic control tasks. In doing so, we investigate different reward formulations - dense, sparse, and non-Markovian. While step-based algorithms perform well only on dense rewards, ERL performs favorably on sparse and non-Markovian rewards. Moreover, our results show that the sparse and the non-Markovian rewards are also often better suited to define the desired behavior, allowing us to obtain considerably higher quality policies compared to step-based RL.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2210.09622")
get_code_for_paper("2210.09622")
have("2210.09622")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2210.09622.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2210.09622.svg)](https://syntology.ai/paper/2210.09622)

Connect an agent — have() is free.